Populate DamageTable from forward scrub
We detect metadata damage two ways: when we try to read things off disk during normal operation, and when we try to verify what's on disk during forward scrub.
Currently DamageTable is only populated in the first path. The issues detected by forward scrub go to the cluster log, but are not visible to the user when they subsequently go to dump the damage table.
Update the forward scrub code to populate DamageTable when it finds errors.